Abstract
The aging population experiences concurrent brain aging and deterioration of bone health. Imaging-derived brain age gap (BAG) demonstrates enhanced predictive capacity for age-related pathologies compared to chronological age. This study included 28,705 participants who underwent brain MRI at a mean age of 63.2 years, with brain age predicted from 1705 imaging-derived phenotypes using LASSO regression (mean predicted brain age: 63.2 years). We then assessed the associations of BAG with BMD at 4 sites and fracture risks. Each 1-year increase in BAG was associated with reduced femoral neck BMD, femoral trochanter BMD, lumbar spine BMD, total body BMD (β(SE) = −0.0028 (0.0003), P = 7.31E − 21; β(SE) = −0.0031 (0.0003), P = 4.04E − 26; β(SE) = −0.0036 (0.0004), P = 1.30E − 16; β(SE) = −0.0033 (0.0002), P = 3.51E − 36, respectively), and increased risks of all-site fractures (HR 1.06, 95% CI: 1.02-1.10). Sex and menopausal status significantly modified the association between BAG and BMD. Findings suggest that higher BAG was associated with lower BMD and higher all-site fracture risk, and these associations may be stronger in men and postmenopausal women.